Household heterogeneity and Incomplete Financial Markets: Asset Return Implications in a Real Business Cycle Setup

Abstract

Uno de los problemas principales de la literature moderna de ciclos es la imposibilidad de replicar el comportamiento de los rendimientos de activos financieros. Varios autores han incorporado mercados financieros en el modelo basico de ciclos, demonstrando que, para cualquier calibracion, este tipo de modelos predice una prima de riesgo de practicamente cero.<br/> Una de las razones del fracaso de estos modelos es que utilizan un agente representativo. En este caso los mercados financieros son completos. En esta tesis relajamos este supuesto al incorporar riesgo idiosincratico, que genera heterogeneidad entre la poblacion. Uno de los objectivos principales es ver si estas extensiones pueden ayudar a mejorar las predicciones del modelo basico con respecto a los rendimientos de activos financieros.<br/> Debemos mencionar que, desde la formulacion original de los puzzles financieros de Mehra y Prescott (85), ha surgido una enorme literatura que intenta analizar las implicaciones de modelos con heterogeneidad respecto a los rendimientos de los activos financieros. Entre otros Aiyagari and Gertler (91), Heaton and Lucas (96), Lucas (94), Marcet and Singleton (99), y Telmer (93) han estudiado este problema bajo el supuesto de que el consumo es exogeno. Si embargo, en nuestro analisis incorporamos production y por tanto ofrecemos un modelo mucho mas apropiado para estudiar los rendimientos financieros. En particular, el consumo se deriva de la maximizacion de la utilidad y por tanto no es exogeno. Ademas, el valor de las acciones se determina a traves de la optimizacion de la empresa.<br/> La presencia de un sector de produccion implica que tenemos que tratar un tema al que no se le ha dado demasiada importancia hasta ahora en la literatura. Con mercados financieros incompletos y heterogeneidad de los accionistas, el problema de maximizacion del valor de la empresa no esta bien definido. Esto significa que tenemos que incorporar objectivos que no son estandares. Una de las contribuciones importantes de este trabajo es demostrar como se puede solucionar este problema.

One of the main problems with the modern real business cycle (RBC) literature is its inability to replicate the empirical behavior of the main asset returns in the data. Several authors have incorporated financial markets into the basic model showing that, regardless of the parameterization or the incorporation of other frictions, like capital adjustment costs, these models are unable to replicate the key financial statistics in the data, predicting an equity premium which is essentially zero, and asset return volatilities that are also far from reality.<br/> One of the main reasons for the lack of success of the previous models may be the fact that they are using a representative agent environment. In this case, financial markets are effectively complete, independently of the existing asset structure. In the present thesis, this assumption is relaxed by incorporating both, idiosyncratic labor income risk and imperfect risk sharing, leading to ex-post household heterogeneity and to an incomplete financial market structure. One of the main objectives is therefore to see, if these extensions can help to improve the asset pricing implications of the standard model.<br/> We have to mention that, since the original statement of the asset pricing puzzles by Mehra and Prescott (85), there has been a large strand of literature trying to analyze the asset pricing implications of a context with household heterogeneity and incomplete financial markets. Among others, Aiyagari and Gertler (91), Heaton and Lucas (96), Lucas (94), Marcet and Singleton (99), and Telmer (93) have studied such a framework under the assumption of exogenously determined asset returns and consumption processes. Note, however, that our analysis goes one step further in the sense that it incorporates a production technology, offering a better foundation of asset prices than the standard exchange economy. In particular, consumption is derived from explicit utility maximization instead of being specified exogenously. In addition, the value of price of equity is determined endogenously via the optimization problem of the firm, which also breaks the identity between dividends and consumption processes in exchange economies. We indeed believe, that a detailed and rigorous analysis of asset pricing requires a general equilibrium model of this type.<br/> Note also that the presence of a non-trivial production sector involves addressing an important issue, which has not been given very much attention in the previous asset pricing literature. Under incomplete financial markets and household (shareholder) heterogeneity, the usual profit maximization of the firm is no longer well defined. Thus, unless one assumes that the firm is myopic, in the sense that it solves a static optimization problem by maximizing period by period profits, one has to incorporate non-standard firm objectives into the model. A second important objective or contribution of the present thesis is therefore to illustrate how to get around the problem of the firm by incorporating a firm objective which is adequate for the case in which financial markets are incomplete.